Optical Coherence Tomography as a diagnostic tool

Optical Coherence Tomography as a diagnostic tool
Ann Singh1,2 Aletta Elizabeth Karsten1
1
Biophotonics Group, National Laser Centre, CSIR, P. O. Box 395, Pretoria, South
Africa, 0001 2 Laser Research Institute, Stellenbosch University, Private bag X1,
Matieland, 7602
E-mail: asingh1@csir.co.za
Abstract. Optical Coherence Tmography (OCT) has been used in biomedical applications as a
method to non-invasively detect changes occurring in tissue such as the detection of skin
cancer The effect of skin tone on detection of skin cancer has however not been studied
extensively. Skin simulating phantoms with different optical properties can be used to optimise
a system before it can be used on patients. For this reason accurately knowing the optical
properties is imperative. An Integrating Sphere system can be used to determine these optical
properties. Preliminary results from phantom and human skin indicate that skin tone may
influence the data obtained from OCT images.
1. Introduction
OCT is a non-invasive imaging technology based on a Michelson interferometer. It is analogous to
ultrasound, however whereas ultrasound uses sound waves, OCT works on light waves, measuring the
backscattering properties of tissues. The light reflected off a sample is interfered with a reference wave
utilizing a technique known as low-coherence interferometry. An interference signal arises when the
optical pathlength difference between the two arms of a Michelson interferometer is within the
coherence length of the source. Hence, the axial resolution in OCT is proportional to the temporal
coherence length of the source. OCT is an imaging technique that fills the gap between confocal
microscopy and ultra-sound imaging. The typical high resolutions offered by OCT are 5−15 µm
although the penetration depth is currently limited to about 2-3 mm. It offers a non-destructive, noncontact, repeatable method of in vivo real time imaging at high speeds [1].
Fercher et al. [2] provide a very concise and nice description of the applications of OCT. “At
present OCT is used in three different fields of optical imaging, in macroscopic imaging of structures
which can be seen by the naked eye or using weak magnifications, in microscopic imaging using
magnifications up to the classical limit of microscopic resolution and in endoscopic imaging, using
low and medium magnification.”
OCT was first demonstrated for retinal imaging [3]. Since then it has been used in biomedical
applications such as ophthalmology, gastroenterology, urology, dentistry, cardiology and dermatology
among other disciplines [4-9]. Atherosclerosis plaques, which contribute to about 75% of
cardiovascular diseases, have been characterised by measuring the backscattering and attenuation
coefficients using OCT [4]. Distinguishing breast tumour through computational methods on OCT
images has also been done [5]. Some other studies include successfully monitoring the changes in
skin after UV radiation [9] as well as in the evaluation of sun damage and pre-cancer diagnosis [10].
Other than skin cancer OCT has been applied in the field of oncology for detection of tumours in the
breast, bladder, brain, gastrointestinal, respiratory, and reproductive tracts, skin, ear, nose, and throat.
In many of these studies it has been able to identify the malignancy in the very early stages which
often is the key to patient survival. OCT has also been explored as an imaging technique for 3-D cell
activities in tissue models with the potential to extend imaging of scaffolds to in vivo applications,
such as following grafting of engineered tissues into a host tissue [11].
Of particular importance to us is the use of OCT in dermatology, where OCT has been used to
visualize the different layers of the skin such as stratum corneum of glabrous skin in palmoplantar
locations, the epidermis and the upper dermis of skin and skin appendages and blood vessels. This has
led to it being a powerful technique for skin cancer detection. Furthermore OCT has been shown to be
useful in non-invasive monitoring of cutaneous inflammation etc.
The aim of our work is to determine how well our commercially bought OCT System (Swept
Source OCT (SS-OCT), Thorlabs) can indentify skin alterations as well as it’s sensitivity to the
influence of skin tone on such detection.
To begin this investigation we will look at a large range of skin simulating phantoms with known
optical properties i.e absorption and reduced scattering coefficients (µa and µ's). The phantoms are
manufactured in-house and measured on an Integrating Sphere (IS) system [12,13]. These phantoms
are then imaged using the OCT system and reconstructed to give a 3D representation of the area
imaged.
2. Methodology
2.1. Preparation of phantoms
Solid resin phantoms are prepared by adding carbon black and TiO particles to a hardener-resin
mixture (Aka-cure,slow, Aka-resin, Akasel) [14,15]. The phantoms are cut into discs of varying
thickness. For these experiments discs of 2mm were used.
2.2. Determination of Optical Properties
The optical properties of the phantoms are determined by using the Integrating Sphere system. In this
particular case, the optical properties at 632.8 nm are determined. The calibration model at this
wavelength has been validated against other models, for a He-Ne laser and a white light source(WL-S)
for single sphere configurations, and shown to have low prediction errors [12,14,15]. Furthermore
currently there are many therapeutic lasers with a red wavelength range. However the optical
properties can be extracted for different wavelengths as we use a supercontiniuum white light source
(Koheras) with wavelength ranging from 500nm to 1750nm. Shown in figure 1 is the IS setup used for
reflectance and transmittance measurements on the phantoms to determine the optical properties. The
details of obtaining the properties are described in [12,13,16].
Figure 1. Single Integrating Sphere setup.
2.3. Taking an image
The system used is a Swept Source OCT (SS-OCT) with a resolution of 10-20 µm. The system is a
frequency domain system with a centre wavelength of 1325nm. The magnitude and time delay of
reflected light is measured and used to construct depth profiles (A-scans). Adjacent A-scans are then
synthesized to create an image. The object to be scanned is placed on the path of the microscope. An
volume of 10x10x3 mm3 (x,y,z) was imaged. The contrast and brightness for all images taken was
kept the same. 3D reconstructions of these images were then made.
2.4. Samples Imaged
Four 2mm phantoms with different optical properties, a 4.2mm phantom with 1mm holes, a standard
pavilion and the index fingers of 2 individuals (two different skin types) were imaged.
2.5. Ethics
Ethics clearance has been obtained for this work. ETHICS REFERENCE NO: N11/01/009.
3. Results
The optical properties of the 4 phantoms were determined using the IS measurements. The
measurements were done using a He-Ne laser and the white light. Table 1 shows the optical properties
predicted at 632.8nm for both set of measurements. The predictions for the WL-S model are
consistently higher than the He-Ne laser.
Table 1: Comparison of the predicted µa and µ's for the solid phantoms using the different models.
Sample
µa He-Ne
µa WL-S
µ's He-Ne
µ's WL-S
A1
0.041
0.05
13.27
14.50
A3
0.11
0.17
10.64
11.90
B1
0.07
0.08
5.47
6.84
B3
0.08
0.12
5.09
6.11
Based on this data these phantoms were shown to have different optical properties. These
phantoms and the other objects were then imaged as described in 2.3 and 2.4. The 3D reconstructions
of these images are shown in Figures (2)-(11).
A
Figure 2. Pavillion.
Figure 3. Phantom with holes.
Figure 4. left index top (person A).
Figure 5. left index top (person B).
Figure 6. left index bottom (person A).
Figure 7. left index bottom (person B).
Figure 8. Phantom A1.
Figure 9. Phantom A3.
Figure 10. Phantom B1.
Figure 11. Phantom B3.
Figure 2 clearly shows the steps on a pavilion. Each step size is 0.5mm. This sample is used as a
quick test to confirm that the system is working properly. Figure 3 shows the 4.2mm phantom with a
hole. Although one can clearly see the hole with the eye, the depth of the hole is not visible to the eye.
However reconstructing the image shows us the depth of the hole (point A). Currently the depth has
not been evaluated, however this image shows us that it may be possible to view and determine the
depth of an object such as a tumour imbedded in skin. Figures 4 and 5 compare the images, taken
from the topside, on the left index finger of A and B. Figures 6 and 7 compare the images, taken from
the bottom, on the left index finger of A and B. Figures 7 and 5 show more clearly defined layers of
the skin than seen in figures 4 and 6. This suggests to us that the skin tone may influence the images
obtained by OCT. However such conclusions can only be made once a study of a larger data set is
done. Furthermore one would need to determine whether the differences are due to skin tone or age,
skin structure etc. The differences in the images obtained for the same person show that the area
imaged also influences the data due to the composition of the skin in the area.
Phantom B1 (Figure 10) appears to show the most detail through the sample and phantom
A3(Figure 9) the least. Phantom B1 has the third lowest absorption and scattering properties whereas
phantom A3 has the highest absorption and second highest scattering properties. Phantom B3 with the
lowest scattering and second highest absorption properties shows more clarity than phantoms A1 and
A3. Finally phantom A1 (Figure 8)with the lowest absorption and highest scattering appears to show a
little more detail than phantom A3. These results are interesting because OCT is based on the
scattering properties of samples. It is the light reflected of the sample that creates the image seen.
However these results show that both the absorption and scattering properties influence the images.
These are encouraging results but need to be re-evaluated with a much larger study.
4. Conclusions
These preliminary findings support our hypothesis that skin tone may influence the OCT images
obtained. It is premature to state that we have proven our hypothesis to be true until the larger study
has been completed.
References
[1] Levitz D, Thrane L, Frosz M, Andersen P, Andersen C, Andersson-Engels S, Valanciunaite J,
Swartling J, and Hansen P 2004 “Determination of optical scattering properties of highlyscattering media in optical coherence tomography images,” Opt. Express 12 249-59
[2] Fercher, AF, Drexler W, Hitzenberger CK and Lasser T 2003 “Optical coherence tomography principles and applications,” Rep. Prog. Phys. 66 239-303.
[3] Huang D, Swanson E A, Lin CP, Schuman JS, Stinson WG, Chang W, Hee MR, Flotte T,
Gregory K, Puliafito CA and Fujimoto JG 1991 “Optical coherence tomography,” Science
254, 1178–81
[4] Faber D, van der Meer F, Aalders M and van Leeuwen T 2004 “Quantitative measurement of
attenuation coefficients of weakly scattering media using optical coherence tomography,”
Opt. Express 12 4353-65
[5] Zysk AM, Nguyen FT, Oldenburg AL, Marks DL and Boppart SA Oct 2007 “Optical coherence
tomography: a review of clinical development from bench to bedside,” J. Biomed. Opt. 12,
051403
[6] Chenyang Xu, Joseph M. Schmitt, Stephane G. Carlier and Renu Virmani, (Jun 2008)
“Characterization of atherosclerosis plaques by measuring both backscattering and
attenuation coefficients in optical coherence tomography,” J. Biomed. Opt. 13(3), 034003
[7] Mujat M, Ferguson RD, Hammer DX, Gittins C and Iftimia N 2009 “Automated algorithm for
breast tissue differentiation in optical coherence tomography,” J. Biomed. Opt. 14, 034040
[8] Mogensen M, Nürnberg B, Forman J, Thomsen J, Thrane L and Jemec G 2009 “In vivo
thickness measurement of basal cell carcinoma and actinic keratosis with optical coherence
tomography and 20-MHz ultrasound,” British Journal of Dermatology 160 1026–33
[9] Gambichler T, Boms S, Stucker M, Moussa G, Kreuter A, Sand M, Sand D, Altmeyer
P and Hoffmann K 2005 “Acute skin alterations following ultraviolet radiation investigated
by optical coherence tomography and histology,” Arch Dermatol Res 297 , pp. 218–25
[10] Korde VR, et al. 2007 “Using optical coherence tomography to evaluate skin sun damage and
precancer,” Lasers Surg Med 39 687–95.
[11] Tan W, Oldenburg AL, Norman JJ, Desai TA and Boppart SA 2006 “Optical coherence
tomography of cell dynamics in three-dimensional tissue models,” Optics Express 14 715971
[12] Singh A, Karsten AE 2011 “Comparison of the accuracy of the calibration model on the double
and single integrating sphere systems,” Proc. 4th International Symposium on
Photoelectronic Detection and Imaging (ISPDI 2011) Beijing China May 24 to 26 2011
[13] Singh A, Karsten AE and Dam JS 2008 “Robustness and accuracy of the calibration model for
the determination of the optical properties of chicken skin,” Proc. of the World Association
of Laser Therapy (WALT), Sun City, South Africa, 19-22 October 4
[14] Karsten AE, Singh A and Braun MW (published online April 2011) “Experimental verification
and validation of a computer model for light–tissue interaction,” Lasers Med Sci, DOI
10.1007/s10103-011-0926-x
[15] Karsten AE and Singh A “Sensitivity of light interaction computer model to the absorption
properties of skin,” Proc. 4th International Symposium on Photoelectronic Detection and
Imaging (ISPDI 2011) Beijing China May 24 to 26 2011
[16] Singh A, Karsten AE, Mputle I, Chetty A and Naidoo K 2009"Determination of the optical
properties of PNIPAAm gels used in biological applications", Proc. SPIE 7373, 737317;
doi:10.1117/12.831882
Acknowledgments
We would like to acknowledge Bafana Moya for preparing the phantoms used in these experiments.